Graduate Diploma in Systems Analysis (GDipSA)
National University of Singapore
- Coursework: Machine Learning, Agent, Web Application Development, Mobile Application Development, DevSecOps, Big Data
Career
Career
Chongqing University undergraduate and GDipSA (NUS-ISS) student combining machine learning with full-stack software engineering — from benchmarked deep-learning models to production systems in Java/Spring Boot/ReactJS
Skills
- Machine Learning
- Deep learning (CNN); gradient boosting & ensemble methods (CatBoost, XGBoost); end-to-end ML workflow with SHAP interpretability
- Programming
- Python, Java, Kotlin, C++, MySQL, R, JavaScript
- Full-Stack Web
- Spring Boot, JPA/Hibernate, ReactJS
- Business Analysis
- ACCA Diploma in Accounting and Business, financial statement analysis, industry research
- Languages
- English (IELTS 7.5, 2025), Chinese (native), Japanese (basic)
Education
Bachelor of Management, Administration Management
Chongqing University
- Designed and executed a supervised ML study predicting internal and external political efficacy using European Social Survey (ESS) Round 11 data covering ~39,000 respondents across 23 European countries
Research
Co-Author
Thesis: Key Predictors of Political Efficacy Based on ML
- Benchmarked 11 models (tree-based ensembles, gradient boosting, SVR, KNN, neural network) and selected CatBoost as the final model based on R², MAE, and RMSE
- Built a reproducible modelling pipeline in Python & R, including complete-case cleaning and categorical encoding
- Applied SHAP interpretability (mean absolute values and dependence plots) to rank feature importance
Projects
Plan-Stage Designer
DevSecOps CI/CD Pipeline Design
- Owned Plan-stage design and coordinated with another team member to align the Plan↔Monitor feedback loop
- Exceeded the Plan-stage scope to help shape team-wide toolchain decisions across containerisation, IaC, and GitOps, and co-designed one end-to-end DevSecOps CI/CD pipeline
Team Leader
Fruit Image Classifier: CNN with Transfer Learning
- Developed a custom PIL-based pre-processing and augmentation pipeline, corrected severe class imbalance (20 vs 75 samples) via weighted sampling and class-weighted loss, and tuned hyperparameters with multi-seed robustness validation
- Led the Version 2 and final optimisation iterations of a 5-person team building a CNN for four-class fruit image classification, applying transfer learning with MobileNetV2 to reach 0.92 accuracy / 0.87 F1
Backend Developer
Full-Stack Shopping-Cart Web Application
- Co-owned backend development with two other teammates, building RESTful API services on a Spring Boot backend with JPA/Hibernate ORM persistence over a SQL database
- Bridged the front-end/back-end boundary by defining API contracts with a 2-person ReactJS frontend team, integrating components into a single e-commerce application spanning 6 core features
Experience
Business Centre Intern
CITIC Securities, Hubei Branch
- Studied extensive internal industry research reports and produced basic industry research reports
- Led a 3-member team to design diversified investment strategies tailored to client needs
- Created 2 ChatGPT-assisted scripts for portfolio-return analysis, applying Python for calculations and Excel VBA for automated reporting, and devised an innovative workflow for the final investment decision
Full write-ups of selected projects and research will gradually land in Wander.